2009 IEEE conference on computer vision and pattern recognition , pages=

representative citing papers

citing papers explorer

Showing 50 of 56 citing papers.

• A Unified Geometric Framework for Weighted Contrastive Learning cs.LG · 2026-05-13 · unverdicted · none · ref 46

  Weighted InfoNCE objectives realize specific target geometries in embedding space, with SupCon producing size-dependent inter-class similarities under imbalance while Soft SupCon and certain continuous variants preserve regular simplex or unique optima.

• Identifying Latent Concepts and Structures for Generalized Category Discovery cs.CV · 2026-07-01 · unverdicted · none · ref 78

  CPF-GCD enforces low-rank compositional structure on vision backbone features via spatial primitive fields so that novel categories emerge as new activation patterns over a shared vocabulary of reusable visual primitives.

• Can Machines Really See Objects in Images? A Study Based on Syntactic Distance and Visual Self-Referential Instances cs.CV · 2026-06-28 · unverdicted · none · ref 8

  Vision models show a phase transition where accuracy on a zero syntactic-distance global-semantics task collapses to chance beyond a critical scale and does not recover with larger models or data.

• Unlearning with Asymmetric Sources: Improved Unlearning-Utility Trade-off with Public Data cs.LG · 2026-05-11 · unverdicted · none · ref 216

  ALU uses public data to suppress unlearning cost quadratically while characterizing distribution mismatch effects, enabling mass unlearning with maintained utility.

• NeuralBench: A Unifying Framework to Benchmark NeuroAI Models cs.LG · 2026-05-08 · conditional · none · ref 299

  NeuralBench is a new benchmarking framework for neuroAI models on EEG data that finds foundation models only marginally outperform task-specific ones while many tasks like cognitive decoding stay highly challenging.

• Empirical Evidence for Simply Connected Decision Regions in Image Classifiers cs.CV · 2026-05-07 · unverdicted · none · ref 19

  Empirical tests with quad-mesh filling indicate that decision regions in modern image classifiers are simply connected.

• Statistical Consistency and Generalization of Contrastive Representation Learning cs.LG · 2026-05-04 · unverdicted · none · ref 75 · 2 links

  The paper proves statistical consistency of contrastive loss to optimal ranking via an AUC criterion and derives generalization bounds O(1/m + 1/sqrt(n)) for supervised and O(1/sqrt(m) + 1/sqrt(n)) for self-supervised CRL that explain benefits of large negative sets.

• VAnim: Rendering-Aware Sparse State Modeling for Structure-Preserving Vector Animation cs.CV · 2026-05-02 · unverdicted · none · ref 27

  VAnim creates open-domain text-to-SVG animations via sparse state updates on a persistent DOM tree, identification-first planning, and rendering-aware RL with a new 134k-example benchmark.

• Frequency-Forcing: From Scaling-as-Time to Soft Frequency Guidance cs.LG · 2026-04-21 · unverdicted · none · ref 26

  Frequency-Forcing guides pixel flow-matching with a data-derived low-frequency auxiliary stream to softly enforce scale-ordered generation, improving FID on ImageNet-256 over baselines.

• One-Step Diffusion with Inverse Residual Fields for Unsupervised Industrial Anomaly Detection cs.CV · 2026-04-20 · unverdicted · none · ref 32

  OSD-IRF performs unsupervised industrial anomaly detection with a single diffusion step by evaluating anomalies in inverse residual field space under a Gaussian, delivering SOTA or competitive results with roughly 2x speedup.

• Learning Interactive Real-World Simulators cs.AI · 2023-10-09 · conditional · none · ref 153

  UniSim learns a universal real-world simulator from orchestrated diverse datasets, enabling zero-shot deployment of policies trained purely in simulation.

• Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference cs.CV · 2023-10-06 · unverdicted · none · ref 45

  Latent Consistency Models enable high-fidelity text-to-image generation in 2-4 steps by directly predicting solutions to the probability flow ODE in latent space, distilled from pre-trained LDMs.

• Diffusion Posterior Sampling for General Noisy Inverse Problems stat.ML · 2022-09-29 · unverdicted · none · ref 59

  Diffusion models solve noisy (non)linear inverse problems via approximated posterior sampling that blends diffusion steps with manifold gradients without strict consistency projection.

• LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale cs.LG · 2022-08-15 · conditional · none · ref 60

  LLM.int8() performs 8-bit inference for transformers up to 175B parameters with no accuracy loss by combining vector-wise quantization for most features with 16-bit mixed-precision handling of systematic outlier dimensions.

• Is the Geometry Doing the Work? An Operating-Point Audit of Hierarchy in Hyperbolic Vision-Language Models cs.CV · 2026-07-06 · conditional · none · ref 38

  Audited MERU, HyCoCLIP, and PHyCLIP checkpoints operate near-Euclidean with saturated entailment cones, so they do not demonstrate active radial or cone-based hierarchy.

• Validating Causal Abstraction Metrics on Simulated Complex Systems cs.LG · 2026-06-30 · unverdicted · none · ref 202

  Authors create a benchmark across discrete/continuous and static/dynamical systems and introduce the Causal Abstraction Error (CAE) metric that reliably distinguishes valid from invalid causal abstractions when it includes faithfulness testing.

• Data Provenance for Image Auto-Regressive Generation cs.CV · 2026-06-22 · unverdicted · none · ref 128

  A post-hoc detection framework exploits generation-induced patterns in autoregressive image outputs to enable provenance tracing across multiple IAR models without altering the generation process.

• Label-Efficient Dataset Pruning via Semi-Supervised Pseudo-Labeling cs.LG · 2026-05-22 · unverdicted · none · ref 65

  SemiPrune uses a small labeled subset and semi-supervised pseudo-labeling to enable supervised dataset pruning methods, achieving state-of-the-art results on domain-specific, image-corrupted, and long-tailed datasets.

• Understanding Generalization through Decision Pattern Shift cs.LG · 2026-05-13 · unverdicted · none · ref 79

  DPS quantifies deviation of per-sample decision patterns from class averages and shows linear correlation with generalization gaps while unifying degradation scenarios into a continuous trajectory.

• Letting the neural code speak: Automated characterization of monkey visual neurons through human language q-bio.NC · 2026-05-12 · unverdicted · none · ref 59 · 2 links

  Natural language descriptions generated via a closed-loop pipeline with digital twins capture the selectivity of most neurons in macaque V1 and V4, with synthesized images driving 96% of V4 neurons into the top or bottom 5% of natural-image response distributions.

• Couple to Control: Joint Initial Noise Design in Diffusion Models cs.LG · 2026-05-11 · unverdicted · none · ref 39

  Coupled initial noises in diffusion models, with designed dependence but unchanged marginal Gaussians, improve generated image diversity on Stable Diffusion variants while preserving quality and alignment.

• LightAVSeg: Lightweight Audio-Visual Segmentation cs.CV · 2026-05-09 · unverdicted · none · ref 37

  LightAVSeg decouples semantic filtering and spatial grounding to achieve linear-cost cross-modal interaction in audio-visual segmentation, reaching 50.4 mIoU on MS3 with 20.5M parameters as a new lightweight state-of-the-art.

• Distributionally Robust Multi-Objective Optimization cs.LG · 2026-05-07 · unverdicted · none · ref 54

  DR-MOO adds distributional robustness to multi-objective optimization and gives single-loop MGDA algorithms reaching epsilon-Pareto-stationary points in O(epsilon^{-4}) samples for nonconvex problems.

• Local Intrinsic Dimension Unveils Hallucinations in Diffusion Models cs.CV · 2026-05-06 · unverdicted · none · ref 32

  Hallucinations in diffusion models are driven by local intrinsic dimension instabilities on the manifold, which Intrinsic Quenching corrects by deflating it.

• On the (In-)Security of the Shuffling Defense in the Transformer Secure Inference cs.CR · 2026-05-06 · conditional · none · ref 110

  An attack aligns differently shuffled intermediate activations from secure Transformer inference queries to recover model weights with low error using roughly one dollar of queries.

• Leveraging Data Symmetries to Select an Optimal Subset of Training Data under Label Noise cs.LG · 2026-05-03 · unverdicted · none · ref 1

  Exploiting data symmetries boosts k-NN to select near-optimal low-noise subsets from noisy datasets, approaching Bayes-optimal performance in high dimensions, with learned representations aiding partial symmetry knowledge.

• Skipping the Zeros in Diffusion Models for Sparse Data Generation cs.LG · 2026-05-03 · unverdicted · none · ref 47

  SED modifies diffusion models to skip zeros during training and inference, preserving sparsity patterns while claiming computational savings and comparable or better generation quality on benchmarks.

• Visual Implicit Autoregressive Modeling cs.CV · 2026-05-02 · unverdicted · none · ref 31

  VIAR embeds an implicit equilibrium layer in next-scale AR image models, reporting FID 2.16 on ImageNet 256 with 38.4% of VAR’s parameters and a per-scale inference compute knob.

• Almost for Free: Crafting Adversarial Examples with Convolutional Image Filters cs.LG · 2026-05-01 · conditional · none · ref 17

  Optimized 3x3 adversarial image filters based on edge detection generate transferable untargeted attacks on neural networks with 30-80% success using only one pass and far fewer parameters than prior methods.

• Possibilistic Predictive Uncertainty for Deep Learning cs.LG · 2026-05-01 · unverdicted · none · ref 14

  DAPPr projects a possibilistic posterior over network parameters to predictions using supremum operators and approximates it with learnable Dirichlet functions to yield an efficient training objective for epistemic uncertainty.

• Parameter-Efficient Adaptation of Pre-Trained Vision Foundation Models for Active and Passive Seismic Data Denoising physics.geo-ph · 2026-04-30 · conditional · none · ref 57

  Adapting vision foundation models with LoRA and kurtosis-guided unsupervised test-time adaptation matches or exceeds domain-specific models for seismic denoising across multiple sites and unseen data.

• Temporal Taskification in Streaming Continual Learning: A Source of Evaluation Instability cs.LG · 2026-04-23 · conditional · none · ref 9

  Different valid temporal partitions of the same streaming dataset can produce materially different rankings and performance numbers for continual learning methods.

• Dual-Anchoring: Addressing State Drift in Vision-Language Navigation cs.CV · 2026-04-19 · conditional · none · ref 49

  Privatar partitions VR avatar reconstruction via frequency-domain decomposition, keeping sensitive components local and offloading the rest with distribution-aware minimal perturbation noise, achieving 2.37x throughput with provable privacy.

• TripoSG: High-Fidelity 3D Shape Synthesis using Large-Scale Rectified Flow Models cs.CV · 2025-02-10 · unverdicted · none · ref 142

  TripoSG generates high-fidelity 3D meshes from input images via a large-scale rectified flow transformer and hybrid-trained 3D VAE on a custom 2-million-sample dataset, claiming state-of-the-art fidelity and generalization.

• Video Prediction Policy: A Generalist Robot Policy with Predictive Visual Representations cs.CV · 2024-12-19 · unverdicted · none · ref 48

  Video Prediction Policy conditions robot action learning on future-frame predictions inside fine-tuned video diffusion models, yielding 18.6% relative gains on Calvin ABC-D and 31.6% higher real-world success rates.

• SalUn: Empowering Machine Unlearning via Gradient-based Weight Saliency in Both Image Classification and Generation cs.LG · 2023-10-19 · conditional · none · ref 155

  SalUn uses gradient-based weight saliency to achieve effective machine unlearning of data, classes, or concepts in image classification and generation, narrowing the gap to exact retraining.

• Multi-Beholder: Biomarker Prediction for Low-Grade Glioma with Multiple Instance Learning and One-Class Classification eess.IV · 2023-10-11 · unverdicted · none · ref 46

  Multi-Beholder integrates one-class classification into multiple instance learning to predict LGG biomarker status from histopathology images, reporting AUCs of 0.973 on TCGA-LGG and 0.820 on an external Xiangya cohort.

• LanguageBind: Extending Video-Language Pretraining to N-modality by Language-based Semantic Alignment cs.CV · 2023-10-03 · unverdicted · none · ref 157

  LanguageBind aligns video, infrared, depth, and audio to a frozen language encoder via contrastive learning on the new VIDAL-10M dataset, extending video-language pretraining to N modalities.

• PixArt-$\alpha$: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis cs.CV · 2023-09-30 · accept · none · ref 156

  PixArt-α matches commercial text-to-image quality with a diffusion transformer trained in 675 A100 GPU days through decomposed training stages, cross-attention text injection, and vision-language model dense captions.

• ST-MoE: Designing Stable and Transferable Sparse Expert Models cs.CL · 2022-02-17 · unverdicted · none · ref 126

  ST-MoE introduces stability techniques for sparse expert models, allowing a 269B-parameter model to achieve state-of-the-art transfer learning results across reasoning, summarization, and QA tasks at the compute cost of a 32B dense model.

• StableGrad: Backward Scale Control without Batch Normalization cs.LG · 2026-05-19 · unverdicted · none · ref 11

  StableGrad applies scale correction to weight gradients after backpropagation to enable stable optimization of deep BatchNorm-free networks including PINNs.

• Temporal Aware Pruning for Efficient Diffusion-based Video Generation cs.CV · 2026-05-18 · unverdicted · none · ref 47 · 2 links

  TAPE applies temporal-aware token pruning with smoothing, reselection, and timestep scheduling to speed up video diffusion models while preserving visual fidelity and coherence.

• Lossless Anti-Distillation Sampling cs.LG · 2026-05-12 · unverdicted · none · ref 63

  LADS is a sampling method that keeps benign user generations statistically identical to the original model while forcing correlated samples across a distiller's multiple accounts, provably worsening their generalization via uniform convergence bounds.

• DIVER:Diving Deeper into Distilled Data via Expressive Semantic Recovery cs.CV · 2026-05-12 · unverdicted · none · ref 27

  DIVER applies a pre-trained diffusion model in a dual-stage process of semantic inheritance, guidance, and fusion to improve semantic expression and cross-architecture generalization in dataset distillation.

• Mutual Enhancement Between Global Tokens and Patch Tokens: From Theory to Practice cs.CV · 2026-05-11 · unverdicted · none · ref 59

  TaTok is a theoretically grounded adaptive tokenization method that uses global tokens and cumulative conditional entropy filtering to reduce redundancy while improving reconstruction quality over fixed-rate patch tokenization.

• APEX: Assumption-free Projection-based Embedding eXamination Metric for Image Quality Assessment cs.CV · 2026-05-08 · unverdicted · none · ref 41

  APEX is an assumption-free image quality metric using Sliced Wasserstein Distance on CLIP and DINOv2 embeddings that claims superior robustness to degradations and cross-dataset stability.

• Region Seeding via Pre-Activation Regularization: A Geometric View of Piecewise Affine Neural Networks cs.LG · 2026-05-07 · unverdicted · none · ref 33 · 2 links

  A pre-activation regularizer seeds more affine regions near data in piecewise affine networks, increasing local region count and improving early training performance.

• Quantifying and Predicting Disagreement in Graded Human Ratings cs.CL · 2026-05-01 · unverdicted · none · ref 31

  Annotation disagreement on toxic language can be moderately predicted from textual features, with high-opposition items proving harder for models to estimate accurately.

• Federated Distillation for Whole Slide Image via Gaussian-Mixture Feature Alignment and Curriculum Integration cs.CV · 2026-05-01 · unverdicted · none · ref 49

  FedHD is a federated learning framework for whole slide images that distills one-to-one synthetic features aligned via Gaussian mixtures and progressively integrates cross-site features through curriculum learning to handle institutional heterogeneity.

• From Codebooks to VLMs: Evaluating Automated Visual Discourse Analysis for Climate Change on Social Media cs.CV · 2026-04-23 · unverdicted · none · ref 27

  VLMs recover reliable population-level trends in climate change visual discourse on social media even when per-image accuracy is only moderate.